Cluster Centers Research Articles

An Accurate Book Spine Detection Network Based on Improved Oriented R-CNN

Book localization is crucial for the development of intelligent book inventory systems, where the high-precision detection of book spines is a critical requirement. However, the varying tilt angles and diverse aspect ratios of books on library shelves often reduce the effectiveness of conventional object detection algorithms. To address these challenges, this study proposes an enhanced oriented R-CNN algorithm for book spine detection. First, we replace the standard 3 × 3 convolutions in ResNet50’s residual blocks with deformable convolutions to enhance the network’s capacity for modeling the geometric deformations of book spines. Additionally, the PAFPN (Path Aggregation Feature Pyramid Network) was integrated into the neck structure to enhance multi-scale feature fusion. To further optimize the anchor box design, we introduce an adaptive initial cluster center selection method for K-median clustering. This allows for a more accurate computation of anchor box aspect ratios that are better aligned with the book spine dataset, enhancing the model’s training performance. We conducted comparison experiments between the proposed model and other state-of-the-art models on the book spine dataset, and the results demonstrate that the proposed approach reaches an mAP of 90.22%, which outperforms the baseline algorithm by 4.47 percentage points. Our method significantly improves detection accuracy, making it highly effective for identifying book spines in real-world library environments.

Machine-Learning-Based Spectral Modeling: A Biomimetic Guide for Enhancing Esthetics.

To provide guidelines and means for optimal coverage and distribution of computer models with 1-10 clusters, designed based on an invivo extensive dental colorimetric database and compare the findings with some reputable shade guides. The Fuzzy C-Means (FCM) algorithm was used to optimize the tab position in the CIELAB color space, while MATLAB Fuzzy Logic was used to implement the codes. A minimization process was performed to fine-tune the cluster centers, minimizing Coverage Error (CE00) and Maximum Error (ME00). Spectral reconstruction was performed, and the results were compared with the corresponding data for VITA classical A1-D4 (VC), Linearguide 3D-Master (LG), and Bleachedguide 3D-Master (BG) shade guide. Paired t-test was employed to assess the significance of CE00 differences. CE00 and ME00 ranges for 1-to-10-tab models were 3.8 (14.2) to 1.7 (4.3), while the corresponding values for shade guides were 3.1 (9.1) for VC, 2.3 (6.0) for LG, and 2.9 (9.3) for BG. Hence, the 4-tab-model with CE00 of 2.2 and ME00 of 7.1, outperformed CE00 of LG (the lowest/best CE00 among shade guides), while the 6-tab-model, with CE00 of 2.0 and ME00 of 5.6, outperformed its ME00. A paired t-test confirmed that increasing the number of tabs models resulted in significant CE improvement in all cases (p ≤ 0.05). Shade guide models with only 4 to 6 tabs, optimized using computational techniques, outperformed physical dental shade guides tested, exhibiting significantly lower (better) Coverage Error and Maximum Error. Given that the spectral modeling was performed on an extensive invivo database of natural teeth, this approach was validated as a biomimetic guide for shade guides and corresponding materials of the future, offering a possibility for better shade matching with fewer shades/shade tabs.

New constraints on the central mass contents of Omega Centauri from combined stellar kinematics and pulsar timing

We performed a combined analysis of stellar kinematics with line-of-sight accelerations of millisecond pulsars (MSPs) to probe the mass content of Omega Centauri (OC ). Our mass model includes the stellar mass distribution, a more concentrated mass component linked to the observed MSP population, a generic cluster of stellar remnants (assumed to be more concentrated than the stars and MSPs), and an intermediate-mass black hole (IMBH), allowing us to determine which of these is statistically preferred to account for these observations. We mass-modeled OC using the package GravSphere to solve the Jeans equations, including constraints in the form of proper motions, line-of-sight velocities, the surface density profile of the stars, the spatial distribution of MSPs, and the recently measured line-of-sight accelerations of a subset of these MSPs, self-consistently modeling their intrinsic spin-down. We explore the impact of different assumed centers of OC on our results and we infer the posterior distributions of the model parameters from the combined likelihood using the nested sampling package dynesty Our analysis favors an extended central mass of $ over an IMBH, setting a 3sigma upper limit on the IMBH mass of $6 We find that pulsar timing observations are an important additional constraint, favoring a central mass distribution that is $ 20<!PCT!>$ more massive and extended than implied by models that are constrained by the stellar kinematics alone. Finally, we find a 3sigma confidence level (CL) upper bound of $6 on the total mass traced by the MSPs, with the density profile following $ p (r) star (r)^ with $ 0.3,$ where $ star (r)$ is the stellar mass density and $ is the stellar velocity dispersion profile. This favors models in which MSPs form via stellar encounters, as in the leading paradigm whereby MSPs are the progeny of low-mass X-ray binaries. Our analysis demonstrates how combining stellar kinematics with MSP accelerations produces new constraints on mass models, shedding light on the presence or absence of IMBHs at the centers of globular clusters. Further, we provide the first validation of its kind where MSP positions are linked to their place of formation in globular clusters, which is in excellent agreement with the expectations of stellar encounter models of MSP formation. This sets a promising precedent amid the rapid growth in the number of observations and discoveries currently taking place in this field.

Antarctic Sea ice distribution detection based on improved ant colony algorithm

The changes in the Antarctic sea ice area are directly related to the changes in the atmosphere and oceans. Determining the Antarctic sea ice distribution is of great significance to the global climate change analysis. The ant colony algorithm adopts a positive feedback mechanism to continuously converge the search process and ultimately approaches the optimal solution, making it easy to find the optimal segmentation threshold for detecting the sea ice distribution. However, the ant colony algorithm has the problems of high computational complexity and easy getting stuck in local optima. In order to better apply the ant colony algorithm to sea ice distribution detection, an improved ant colony algorithm was proposed, which improves the selection of initial clustering centers and the update of pheromone volatilization factors in the ant colony algorithm. We compared the improved ant colony algorithm with iterative algorithm, maximum entropy algorithm, and basic global threshold algorithm, and the results showed that the proposed algorithm is feasible. To further validate the accuracy of the improved ant colony algorithm, we compared the results obtained from MODIS data with the improved ant colony algorithm, iterative algorithm, maximum entropy algorithm, and basic global threshold algorithm for sea ice detection, and the results showed that the accuracy of the proposed algorithm was 4.99%, 3.66%, and 5.46% higher than the other three algorithms, respectively.

Granularity Optimization of Travel Trajectory Based on Node2vec: A Case Study on Urban Travel Time Prediction

Intersections are known to cause significant changes in traffic states. However, existing link-level trajectory optimization methods often overlook intersection information, making it challenging to preserve key traffic state features during the optimization process. To address this limitation, a novel approach is proposed that integrates node2vec and K-means algorithms. First, the role of intersections in linking road segments is considered. The node2vec algorithm is employed to capture the deep spatial similarity between links while weakening the adjacency relationship between links before and after intersections. This process generates feature representations for each link. Subsequently, clustering centers are initialized at the intersections, and K-means clustering is applied based on these link feature representations. Through this method, consecutive links within a trajectory that belong to the same cluster are merged, thus optimizing the granularity of the trajectory. Finally, experimental analysis and validation are conducted using link-level travel trajectory data from Shenzhen. The results demonstrate that, under optimal conditions, the mean absolute error (MAE), the mean absolute percentage error (MAPE), and the root mean square error (RMSE) values are reduced by 8.91%, 9.44%, and 8.96%, respectively, while computational efficiency is increased by 30.08%. The proposed trajectory granularity optimization method, which accounts for the existence of intersections, not only effectively retains the key traffic state features from the original trajectory but also significantly reduces training time while improving the model’s prediction accuracy.

Unsupervised contrastive clustering via density cluster representative combination

The recent development of contrastive clustering for deep image clustering has shown promising results by combining representation learning and clustering prediction into a unified framework. In this paper, we introduce an unsupervised contrastive clustering method inspired by contrastive learning to address the issue of cluster center selection in density peaks clustering. We propose a new type of information entropy peaks based on dynamic reverse nearest neighbor sequence to measure local density for each object, and select a small number of top information entropy peaks as cluster representative candidates to cover the right cluster centers as comprehensively as possible. To mitigate the risk of wrongly determining cluster centers, we generate a cluster centers set by computing various cluster representative candidates combinations. In the contrastive clustering stage, we introduce transitive nearest neighbor distance to compute cluster-level loss for the core data set. The prototype clustering is obtained with the minimum cluster-level loss, and the non-core data set is grouped into prototype clustering by nearest neighbor distance. Our comprehensive experiments on both artificial and real-world data sets demonstrate that the proposed unsupervised contrastive clustering method is effective and outperforms the compared state-of-the-art clustering algorithms.

Jet Interaction with Galaxy Cluster Mergers

Active galactic nucleus (AGN) bubbles in cool-core galaxy clusters are believed to facilitate the transport of cosmic-ray electrons (CRe) throughout the cluster. Recent radio observations reveal the complex morphologies of cluster diffuse emission, potentially linked to interactions between AGN bursts and the cluster environment. We perform 3D magnetohydrodynamical simulations of binary cluster mergers and inject a bidirectional jet at the center of the main cluster. Kinetic, thermal, magnetic, and cosmic ray (CR) energy are included in the jet and we use the two-fluid formalism to model the CR component. We explore a wide range of cluster merger and jet parameters. We discuss the formation of various wide-angle-tail and X-shaped sources in the early evolution of the jet and merger. During the last phase of the evolution, we find that the CR material efficiently permeates the central region of the cluster reaching radii of ∼1–2 Mpc within ∼5–6 Gyr, depending on the merger mass ratio. We find that solenoidal turbulence dominates during the binary merger and we explore the possibility for the CR jet material to be reaccelerated by super-Alfvènic turbulence and contribute to cluster scale radio emission. We find high volume fractions, ≳70%, at which the turbulent acceleration time is shorter than the electron cooling time. Finally, we study the merger shock interaction with the CRe material and show that it is unlikely that this material significantly contributes to the radio relic emission associated with the shocks. We suggest that multiple jet outbursts and/or off-center radio galaxies would increase the likelihood of detecting these merger shocks in the radio due to shock reacceleration.

Expansion kinematics of young clusters

Context. Most stars form in clusters or associations, but only a small number of these groups are expected to remain bound for longer than a few megayears. Once star formation has ended and the molecular gas around young stellar objects has been expelled via feedback processes, most initially bound young clusters lose the majority of their binding mass and begin to disperse into the Galactic field. Aims. This process can be investigated by analysing the structure and kinematic trends in nearby young clusters, particularly by analyzing the trend of expansion, which is a tell-tale sign that a cluster is no longer gravitationally bound and dispersing into the field. Methods. We combined Gaia DR3 five-parameter astrometry with calibrated RVs for members of the nearby young cluster λ Ori (Collinder 69). Results. We characterised the plane-of-sky substructure of the cluster using the Q-parameter and the angular dispersion parameter. We find evidence that the cluster contains a significant substructure but that this is preferentially located away from the central cluster core, which is smooth and likely remains bound. We found strong evidence for expansion in λ Ori in the plane of sky by using a number of metrics, but we also found that the trends are asymmetric at the 5σ significance level, with the maximum rate of expansion being directed nearly parallel to the Galactic plane. We subsequently inverted the maximum rate of expansion of 0.144−0.003+0.003 kms−1 pc−1 to give an expansion timescale of 6.944−0.142+0.148 Myr, which is slightly larger than the typical literature age estimates for the cluster. We also found asymmetry in the velocity dispersion as well as signatures of cluster rotation, and we calculated the kinematic ages for individual cluster members by tracing their motion back in time to their closest approach to the cluster centre.

Confirming the Tidal Tails of the Young Open Cluster Blanco 1 with TESS Rotation Periods

Blanco 1 is an ≈130 Myr open cluster located 240 pc from the Sun, below the Galactic plane. Recent studies have reported the existence of diffuse tidal tails extending 50–60 pc from the cluster center based on the positions and velocities measured by Gaia. To independently assess the reality and extent of this structure, we used light curves generated from TESS full-frame images to search for photometric rotation periods of stars in and around Blanco 1. We detected rotation periods down to a stellar effective temperature of ≈3100 K in 347 of the 603 cluster member candidates for which we have light curves. For cluster members in the core and candidate members in the tidal tails, both within a temperature range of 4400–6200 K, 74% and 72% of the rotation periods, respectively, are consistent with the single-star gyrochronological sequence. In contrast, a comparison sample of field stars yielded gyrochrone-consistent rotation periods for only 8.5% of the stars. The tidal tail candidates’ overall conformance to the core members’ gyrochrone sequence implies that their contamination ratio is consistent with zero and <0.33 at the 2σ level. This result confirms the existence of Blanco 1 tidal tails and doubles the number of Blanco 1 members for which there are both spatio-kinematic and rotation-based cluster membership verification. Extending the strategy of using TESS light curves for gyrochronology to other nearby young open clusters and stellar associations may provide a viable strategy for mapping out their dissolution and broadening the search for young exoplanets.

A novel two-stage energy sharing method for data center cluster considering ‘Carbon-Green Certificate’ coupling mechanism

The energy consumption of data centers is on a sharp upward trend, and how to achieve joint optimal scheduling and energy management of multiple data centers is the key to the economic and low-carbon operation of data center prosumers (DCPs). This study introduces shared electric-hydrogen energy storage (SEHES) for DCPs, constructs a data center cluster (DCC) framework, and proposes a two-stage energy sharing model for DCC considering the ‘Carbon-Green Certificate’ coupling mechanism (C-GCCM). First, the DCC framework containing DCC operator (DCCO) and DCPs is established, and the C-GCCM based on conversion coefficients is designed to improve the energy saving and emission reduction level of DCC. In the day-ahead optimization stage, a stochastic multi-objective scheduling model for SEHES considering economy and environmental protection is developed. In the real-time optimization stage, considering the C-GCCM and the difference in on-grid prices, the Stackelberg game model of DCC is constructed according to the internal energy price set by DCCO and the trading strategies of DCPs. Finally, the proposed model is solved using the normal boundary intersection algorithm, the particle swarm algorithm and CPLEX. Simulation results demonstrate that the proposed DCC framework can reduce the daily operating cost of DCPs by 67.05%. The introduction of C-GCCM has resulted in a shift from negative to positive benefits on carbon emission trading for DCC and a 42.33% increase in benefits on green certificate trading.

Pengelompokkan daerah rawan bencana di kabupaten Boyolali menggunakan algoritma k-means

The country of Indonesia, which is located in the Southeast Asia region, faces a high challenge of disasters due to demographic, geological and geographical factors that influence the occurrence of disasters, both those caused by natural, non-natural and human factors. In the Province of Central Java, especially Boyolali, disasters pose a significant threat, especially when the weather is uncertain. However, to identify priority areas for the lowest and highest levels of natural disasters in Boyolali, it is necessary to improve using data mining methods, therefore to facilitate analysis and grouping of data in identifying disaster-prone areas in Boyolali, the K-Means clustering algorithm is used. The method used in this research is qualitative and the data is analyzed using the Rapidminer Studio Ver application. 9.10. Based on the results of the analysis, there is a division of disaster-prone areas into 3 clusters in Boyolali Regency, namely very high vulnerability, medium vulnerability, and low vulnerability. The results of calculations with Rapidminer show that there are 3 sub-districts in cluster 2, 2 sub-districts in cluster 1, and 17 sub-districts in cluster 0. It should be noted that manual calculations can produce differences because the cluster center point or centroid is taken randomly for each calculation method. It is hoped that this grouping of areas can be used as a support for the Boyolali regional BPBD to focus on anticipatory steps to reduce the impact of disasters on society that may occur in the future.

Analyzing the Efficacy of K-Means Clustering and Logistic Regression For Diabetes Prediction

Diabetes causes a large number of deaths each year and a large number of people living with the disease do not realize their health condition early enough. In this study, we propose a data mining based model for early diagnosis and prediction of diabetes using the UCI database. Although K-means is simple and can be used for a wide variety of data types, it is quite sensitive to initial positions of cluster centers which determine the final cluster result, which either provides a sufficient and efficiently clustered dataset for the logistic regression model, or gives a lesser amount of data as a result of incorrect clustering of the original dataset, thereby limiting the performance of the logistic regression model. Our findings offer insights into the comparative strengths and weaknesses of each method, shedding light on their potential applications in diabetes diagnosis and risk assessment. A further experiment with a new dataset showed the applicability of our model for the predication of diabetes.

Origin of high dark remnant fractions in Milky Way Globular Clusters: The crucial role of initial black hole retention

Abstract Comparing the dynamical and stellar masses of Milky Way (MW) globular clusters (GCs) reveals a discrepancy exceeding a factor of two. Since this substantial invisible mass is concentrated in the cluster centre, it is attributed to stellar remnants. The majority of mass in remnants consists of white dwarfs (WDs). Allocating over half of a GC’s current mass to WDs could significantly restrict the dynamical evolution scenarios governing stellar clusters. As the most massive stars in GCs, black holes (BHs) exert a substantial effect on the escape rate of lower mass stars, such as WDs. This paper aims to identify which scenarios of BH natal kicks can accurately reproduce the notable dark remnant fraction observed in MW GCs. We compare the observed remnant fraction of MW GCs with a comprehensive grid of direct N-body simulations while adjusting the natal kick received by BHs. Our results reveal that simulations employing low natal kicks to BHs are the only ones capable of mirroring the remnant fraction of MW GCs. According to the Spitzer instability, the presence of a BH population prompts the formation of a BH sub-system (BHSub) at the centre of a star cluster. The BHSub serves as an energetic power plant, continually releasing kinetic energy through few-body encounters between single and binary BHs, and transferring the generated energy to the entire stellar population. This energy induces a significant difference in the ejection rate of stellar remnants and luminous stars, ultimately increasing the fraction of dark remnants within the star cluster.

Probing the High-frequency Variability of NGC 5044: The Key to Active Galactic Nucleus Feedback

The active galactic nucleus (AGN) feeding and feedback process in the centers of galaxy clusters and groups is still not well understood. NGC 5044 is the ideal system in which to study AGN feedback. It hosts the largest known reservoir of cold gas in any cool-core galaxy group, and features several past epochs of AGN feedback imprinted as cavities in the X-ray-bright intragroup medium, as well as parsec-scale jets. We present Submillimeter Array, Karl G. Jansky Very Large Array, James Clerk Maxwell Telescope, and Atacama Large Millimeter/submillimeter Array high-frequency observations of NGC 5044 to assess the time variability of the millimeter-wave band emission from the accretion disk, and quantify the spectral energy distribution (SED) from the radio to submillimeter band. The SED is well described by an advection-dominated accretion flow (ADAF) model and self-absorbed jet emission from an aging plasma with τ ∼ 1 kyr. We find a characteristic variability timescale of 150 days, which constrains the ADAF emission region to about 0.1 pc, and the magnetic field to ∼4.7 mG in the jets and 870 G in the accretion disk. Longer monitoring/sampling will allow to understand if the underlying process is truly periodic in nature.

A Review on the Application of Automatic Text Summarization and Its Bias Analysis

Abstract. In the era of information explosion, the Internet produces a large number of text documents daily, providing rich information but also posing a challenge: how to swiftly extract key information. Traditional manual reading is time-consuming and inadequate for handling vast data. Thus, automatic text summarization technology emerges as a crucial solution. This paper reviews the application and deviation analysis of this technology across various fields, focusing on addressing shortcomings of traditional methods, such as initial cluster center selection and redundancy. An automatic text summarization method based on an improved TextRank algorithm and K-Means clustering is introduced. Existing methods often struggle with inaccurate initial clustering center selection and high summary redundancy, especially with long texts, resulting in summaries that fail to reflect core content accurately. Furthermore, the widespread use of pre-trained language models introduces potential biases that can propagate to downstream tasks, affecting summary accuracy and impartiality. To address these issues, this paper proposes an innovative automatic text summarization method that optimizes initial clustering center selection and clustering refinement strategies to enhance summary accuracy and readability. Additionally, it discusses name-nationality bias in pre-trained language models and its propagation in text summary tasks, offering a theoretical foundation and practical guidance for developing a more just and reliable Natural Language Processing (NLP) system.

Evaluation method of distribution network operation status based on local fuzzy measure in boundary region

With the increasing complexity of the distribution network, the proportion of abnormal data in the monitoring data of the distribution network and its daily work is extremely low. Traditional clustering analysis methods are difficult to effectively solve the imbalance problem. Therefore, this paper introduces the influence parameters that can adaptively adjust the cluster center of local samples in the boundary area, and improves the cluster center update formula, and proposes a method of distribution network operation state evaluation based on the local blur measurement of the boundary region. The research results found that the five evaluation indicators of the proposed algorithm were 112, 0, 2, 26, and 5, respectively, all of which were superior to the comparison algorithms. The research results showed that the cluster center update optimization method based on local fuzzy measure in boundary region could effectively reduce the negative impact of the edge region occupied by most clusters on its clustering effect, so that the cluster center was always in an ideal position. At the same time, the example results showed that the research method had a risk prediction of 0.91 for power outage networks, which was close to the real situation and had high accuracy. It can provide reference for the operation and maintenance work of power grid personnel, eliminate hidden dangers in advance, and ensure the safe operation of the power grid.

Control parameter optimization of dry hobbing under user evaluation

The appropriate selection of dry hobbing control parameters can significantly improve cutting performances such as energy consumption and time. To overcome the problems of manually setting the value range of control parameters and numerical oscillation of the optimization results (i.e. low stability), a hybrid optimization approach is proposed for the improvement of hobbing performances under user evaluation using K-means clustering, multi-objective hunger games search (MOHGS) and TOPSIS. Firstly, K-means clustering is applied to obtain the cluster centers and case clusters based on the historical cutting cases. Considering the problem of control parameters to be optimized, the numerical range of control parameters is determined via the cluster centers and case clusters. Then, MOHGS is developed and applied to search the non-inferior control parameters and run for many times based on the multi-target optimization model of dry hobbing. Finally, all the non-inferior control parameters obtained are ranked by TOPSIS method based on the user evaluation of cutting performances, and the control parameter ranking first is for machining. Compared with other well-known approaches, it is proved that the proposed approach can effectively resolve the control parameter optimization problem of dry hobbing under user evaluation and has high stability.

A Young Super Star Cluster Powering a Nebula of Retained Massive Star Ejecta

We suggest that “Godzilla” of the lensed Sunburst galaxy (z = 2.37) is a young super star cluster powering a nebula of gravitationally trapped stellar ejecta. Employing Hubble Space Telescope photometry and spectroscopy from the Very Large Telescope (VLT) MUSE and VLT/X-Shooter, we infer the physical and chemical properties of the cluster and nebula, finding that Godzilla is young, 4–6 Myr; massive, 2 × 106 M ⊙ (1000/μ); of stellar metallicity, Z ≃ 0.25 Z ⊙; and has a compact far-UV component of ≲1 pc (1000/μ), where μ is the flux magnification factor. The gas is significantly enriched with N and He, indicating stellar wind material, and has highly elevated O relative to the stellar metallicity, indicating entrainment of core-collapse supernova (CCSN) ejecta. The high density, n e ≃ 107−8 cm−3, implies a highly pressurized intracluster environment. We propose that the pressure results from CCSN-driven supersonic turbulence in warm, self-shielding gas, which has accumulated in the cluster center after runaway radiative cooling and is dense enough to resist removal by CCSNe. The nebula gas shows subsolar C/O, Ne/O, and Si/O, which may reflect the CCSN element yields for initial stellar masses >40 M ⊙. A comparison to element yield synthesis models for young star clusters shows that the gas abundances are consistent with complete retention and mixture of stellar winds and CCSN ejecta until the inferred cluster age. The inferred O and He enhancement may have implications for the formation of multiple stellar populations in globular clusters, as stars formed from this gas would contradict the observed abundances of second-population stars.

Mixed Carboxylate Ligands Bridging Tetra-Pr3+-Encapsulated Antimonotungstate: Syntheses, Structure, and Catalytic Activity for Imidazoles Synthesis.

Multinuclear Pr-containing antimonotungstate [Pr4(H2O)10W6O13(mal)2(OAc)(B-α-SbW9O33)4]21- (Pr-1, mal = malate anion, OAc = acetate anion), bridged by organic carboxylic acid, was synthesized through a one-pot assembly reaction and structurally characterized. Pr-1 is composed of four [B-α-SbW9O33]9- fragments fused together by an organic-inorganic hybrid central [Pr4(H2O)10W6O13(mal)2(OAc)]15+ cluster core through 24 μ2-O atoms. Notably, the central cluster comprises unprecedented decanuclear Pr4(H2O)10W6O13 jointly decorated by two types of carboxylic acid ligands. This integration of rare earth-containing antimonotungstate with mixed organic carboxylate ligands is very rare in POMs chemistry. Pr-1 exhibits excellent catalytic activity in the cyclo-condensation reaction involving benzil, aldehyde, and NH4OAc. A series of 2,4,5-trisubstituted imidazoles were synthesized in remarkable yields using iPrOH as a green solvent.

Evaluation Method of Water Injection Development Effects in Fractured Vuggy Carbonate Reservoirs: Case Study of FI7 Strike-Slip Fault Zone

As the main method of secondary development, water injection has been widely used in fractured vuggy carbonate reservoirs. By using an evaluation index system combined with an integrated approach to objectively evaluate and support targeted and operational adjustments of water injection development effects in fractured vuggy carbonate reservoirs, a comprehensive evaluation method is established in this study. The CRITIC method is used as the main approach, and the analytic hierarchy process, the entropy weight method, and the coefficient of variation method are used as sub-methods. Additionally, the clustering centres are divided using the clustering method to reduce the error caused by the irrational distribution of the relevant evaluation index data. The method is used to evaluate and analyse the effects of water injection development, using the FI7 fault zone of the Halahatang oilfield as an example. The application results show that the evaluation method is feasible and effective for cases with small data volume, fewer computational resources, and less time. This study has a certain reference significance for the evaluation of the effect of water injection development in similar fractured vuggy reservoirs.